analysis of trends and variations of monthly mean wind speed data in nigeria

IOSR Journal of Applied Physics (IOSR-JAP)

e-ISSN: 2278-4861.Volume 7, Issue 4 Ver. I (Jul. - Aug. 2015), PP 31-41

www.iosrjournals.org

DOI: 10.9790/4861-07413141 www.iosrjournals.org 31 | Page

Analysis of Trends and Variations of Monthly Mean Wind Speed

Data in Nigeria

Solomon Okechukwu Amadi 1*

And Sunday Okon Udo 2

1 Dept Of Physics, Geology And Geophysics, Federal University Ndufu-Alike Ikwo,

2dept Of Physics, University Of Calabar,

Abstract: Trends and variations of monthly mean wind speed data in Nigeria were analyzed. The data used were obtained from the Nigerian Meteorological Agency, Oshodi, Lagos. 20 land anemometer stations across

various ecological zones and climatic belts in Nigeria were selected for the analyses. The data length spanned

from 19512012 with some variations in data length across the stations. Statistical techniques used for the analyses are Mann-Kendalls rank correlation tests, simple linear regression, Pearsons product moment correlations, time series plots, descriptive statistics and bar charts. The Mann-Kendalls test results indicate dominant declining trends over the period. 11 stations show downward trends, with 8 showing significant

downward trends at the 1% level. 9 stations show upward trends, with 7 showing significant upward trends at

the 1% level. The Pearsons product moment correlation coefficients indicate that some of the station pairs have negative correlations significant at the 1% and 5% levels. Other station pairs show significant positive

correlations at the 1% and 5% levels while other station pairs show negative and positive correlations that are

not significant at the chosen significance levels. The seasonal variations represented in the bar charts indicate

that spring is the windiest period in most of the stations while autumn dominates the calm period in most

stations across Nigeria. Majority of the stations show high coefficients of variation which increases northwards

along with the monthly mean wind speeds. The results have implications for air quality management, modeling

of wind speed regimes, planning and financing of wind energy and heat and moisture transfer between the

earths surface and the atmosphere.

Key words: Trends, Variations, Wind speeds, Mann-Kendall, Nigeria, Linear Regression.

I. Introduction This research is undertaken to analyze the trends and variations of monthly mean wind speed data in

Nigeria as an index of climate change using Mann-Kendall rank correlation tests and other statistical techniques.

Even though most of the climate change and variability studies have so far focused on temperature and

precipitation, variation in wind speed distribution is also important with respect to the impacts of climate

variability and change (Abhistek et al,2010; Turkes, 1996; Turkes, 1999; Turkes et al, 2008;Zhihua et al,2013;

Amadi et al, 2014;Abiodun et al, 2011; Karabulut et al, 2008; Karaburun et al, 2012;).Tuller (2004) observed

that most practical effects of variations and trends in climate do not involve a single climate parameter but are

the synergistic result of multiple climatic parameters. We have reached a stage in the atmospheric variation

where much effort is focused on other parameters such as wind speed. Different weather systems characterize

different seasons, bringing about a marked seasonal variation in prevailing wind speed and direction.

Almost every impact of climate variation involves wind speed either directly or indirectly (Abhishek et

al, 2010; Tuller, 2004). For instance, one of the ways that air temperature variations affect objects and living

organisms is through sensible heat flux density, which is a function of wind speed. According to Troccoli et al,

(2012), accurate estimates of long-term linear trends of wind speed provide a useful indicator for circulation

changes in the atmosphere and are invaluable for the planning and financing of wind energy.

Until recently, air pollution was thought to be just a problem of the vicinity or locality of occurrence.

New data reveal that air pollution is transported across continents and ocean basins due to fast long-range

transport, resulting in trans-oceanic and trans-continental plumes of atmospheric brown clouds (ABCs)

containing sub micron sized particles called aerosols (Ramanathan and Feng, 2009). Wind is instrumental to the

transport of particulates from industries and mobile sources (Cabezudo et al, 1997), and in the transfer of heat

and moisture between the earths surface and the atmosphere. It therefore follows that the heat and moisture transfer between the earths surface and the atmosphere is attenuated if wind speed decreases over the period. Studies have found that weaker winds in a warmer climate led to higher concentrations in pollution plumes

(Ramanathan and Feng, 2009; Jacob and Winner, 2009; Holzer and Boer, 2001). Reduced wind speed can imply

poor ventilation of pollutants and thereby exacerbating lung and heart diseases especially for asthmatics

(Abhishek et al, 2010). Jacob and Winner (2009) noted that the two air pollutants of most concern for public

health are surface ozone and particulate matter, which are subject to long-range transport by the winds.

According to Hwang et al, (2007), strong correlation exists between ozone distribution pattern, and local and

Analysis Of Trends And Variations Of Monthly Mean...


synoptic meteorological conditions, especially wind speed. The analysis of wind speed patterns is also important

in estimating the surface energy balance (Rayner, 2007) and mitigating coastal erosion (Viles and Goudie,

2003). Wind pattern information is beneficial to agricultural industry (O Neal et al, 2005), and forest and infrastructure protection communities (Jungo et al, 2002). Wind trend analysis is equally important for basic

climatic processes such as evapo-transpiration and land surface atmosphere feedback processes, and also for diverse applications such as wind power generation (Mc Vicar and Roderick, 2010). Furthermore, wind speed

and direction data are useful in air dispersion modeling and identifying pollutant emission sources (Droppo and

Napier, 2008; Wu et al, 2008).

Thus, wind speed is an important element in the study of atmospheric variations, hence the justification

of this paper. Studies on measured wind speed variations have been carried out (Tuller, 2004; Abhishek et al,

2010; Bichet et al, 2012; Ko et al, 2010b; Ewona and Udo, 2008; Mc Vicar et al, 2010; Troccoli et al, 2012).

These studies observed decrease in annual wind speed in numerous sites around the globe during the past few

decades. Roderick et al, (2007) had shown that these observations are not products of measurement artefacts.

Other studies (e.g Kumar and Philip, 2010; Wu and Mok, 2013; Ko et al, 2010a) show that the

direction of trend and variation are location dependent. On the other hand, Cardone et al (1990) observed

surface wind strengthening for marine wind data.

The purpose of this study is to:

1. Examine the trends and variations in measured wind speed data at 20 anemometer stations across Nigeria. 2. Quantify the spatial and temporal relationships of the wind speed data by carrying out correlation analysis

of the individual stations.

3. Examine descriptive statistical features of mean monthly wind speed data of the stations from 1950 2012.

II. Study Area Nigeria co-ordinates on latitude 10.00

oN and longitude 8.00

oE. The climate is tropical; humid in the

south and semi-arid in the north. It comprises various ecotypes and climatic zones. There are two main seasons,

namely, rainy and dry seasons. The rainy season lasts from March to November in the south and May to October

in the north. During December to March, the Nigerian climate is entirely dominated by the north east trade

winds, locally called "harmattan, which originate from Sub-Tropical Anticyclones (STA).Thisharmattan is associated with the occurrence of thick dust haze and early morning fog and mist as a result of radiation cooling

at night under clear skies. The climate is dominated by the influence of Tropical Maritime (TM) air mass, the

Tropical Continental (TC) air mass and the Equatorial Easterlies (EE) (Ojo, 1977) in (Abiodun et al, 2011).

According to Abiodun et al (2011), the TM air mass originates from the southern high-pressure belt located off

the Namibian coast. This air mass becomes a moisture laden air mass after picking up moisture from over the Atlantic Ocean. The TC air mass originates from the high-pressure belt north of the Tropic of cancer. This air

mass is always dry and travels towards Nigeria over the Sahara desert. The TM and TC air masses meet at the

Inter-Tropical Convergence Zone (ITCZ). The EE air mass is an erratic cool air mass which comes from the east

and flows in the upper atmosphere along the ITCZ. The seasonal north-south migration of the ITCZ dictates the

Nigerian weather pattern. Fig 1 is the map of Nigeria indicating the anemometer stations used in the study.

Fig. 1: Map Nigeria showing meteorological locations for the study



III. The Data 3.1 The Database

Monthly mean wind speed data of 20 anemometer stations spread across Nigeria were obtained from

the archives of the Nigerian Meteorological Agency (NIMET) Oshodi, Lagos Nigeria. The period of the data

spanned from 1951 to 2012. Table 1 below gives the summary information of the stations used in the study.

Table 1: Anemometer stations and details of the data used S/N Station Name Latitude

(oN)

Longitude

(oE)

Altitude

(m)

Period Sequence length

(months)

Missing data

(%)

1. Yelwa 10.53 4.45 244 1962-2012 612 11.76

2. Sokoto 12.55 5.12 351 1968-2012 552 2.17

3. Kaduna 10.42 7.19 645 1967-2012 564 10.64

4. Kano 12.03 8.32 476 1961-2012 624 7.69

5. Bauchi 10.17 9.49 591 1961-2012 624 21.15

6. Maiduguri 11.51 13.05 354 1961-2012 624 9.62

7. Ilorin 8.26 4.30 308 1961-2012 624 5.77

8. Yola 9.16 12.26 191 1963-2012 600 2

9. Ikeja 6.35 3.20 40 1951-2012 744 0

10. Ibadan 7.22 3.59 234 1961-2012 624 9.62

11. Oshogbo 7.47 4.29 305 1961-2012 624 11.54

12. Benin 6.19 5.36 77.80 1967-2012 552 0

13. Warri 5.31 5.44 6.00 1967-2012 552 30.43

14. Lokoja 7.48 6.44 113 1964-2012 588 2.04

15. Port Harcourt 5.01 6.57 18 1960-2012 636 0

16. Owerri 5.25 7.13 91 1977-2012 432 0

17. Enugu 6.28 7.34 142 1961-2012 624 5.77

18. Calabar 4.58 8.21 62 1961-2012 624 0

19. Makurdi 7.42 8.37 113 1961-2012 624 13.46

20. Ogoja 6.40 8.48 117 1978-2012 420 17.14

3.2 Data Quality Check And Database Construction

The monthly data required careful scrutiny. This necessitated the construction of the database. Some

missing entries were observed (see table 1 above) and were not replaced. Only 5 stations did not have missing

observations. The missing observations ranged from 2% to about 30%. Shongwe et al (2006) suggested the use

of data from stations with missing records not greater than 5%. However, this can be a major challenge to

achieve especially in data scarce regions (as is the case here). Ngongondo et al (2011) adopted a more flexible

10% maximum threshold recommended by Hosking and Wallis (1997). In the case here, only 13 out of the 20

stations meet the 10% maximum threshold recommendation. Helsel and Hirsch (1992) in National Nonpoint

Source Monitoring Programme (NNSMP) (2011) recommended that monotonic trend analysis could be applied

if the data gap does not exceed one-third of the total record. This recommendation is based on the use of non-

parametric tests that are robust against large data gaps. This is adopted in this study.

A preliminary step in analysis of homogeneity is to plot the time series on a linear scale. Visual

inspection of the plots could reveal the existence of the marked changes in the time series which can be further

investigated by statistical procedures. In the case here, plot inspection immediately revealed the existence of

missing data for some of the stations. Statistical test for homogeneity was done using the non-parametric

Kruskal-Wallis (K-W) test (Turkes et al 2008).

Homogeneity means that there are no jumps (non-climatological abrupt rises or falls) in the climatic

series of observation (Turkes 1999). Most of the statistical in homogeneities noticed in the result of the K-W test

are very much likely related to the long period fluctuations and trends. According to Syners (1990), in Turkes

(1999) and Turkes (1996), these are acceptable within non-randomness characteristics of series of climatological

observations.

IV. Methodology 4.1 Processing Software Packages

Database construction and quality control of the wind speed data were first performed by checking for

missing entries, outliers and temporal homogeneity as discussed in the preceding section. The descriptive

statistics of the distribution was evaluated using the SPSS package. SPSS computer package was also used to

evaluate the Mann-Kendalls rank correlation tests and the Pearsons product moment correlation coefficients. The non-parametric Mann-Kendalls rank correlation tests were used to detect the presence, direction and significance of the trends. The Pearsons product moment correlation coefficients revealed the spatial and temporal relationships of the wind speed data. The Mann-Kendall tests detect trends but cannot provide an

estimate of the trend magnitude. The trend magnitudes were quantified by a linear regression model. Regression

analysis was executed using the MATLAB software package. The Time series plots with the trend lines were



done using the MATLAB. The R programming language was used to do the bar charts to indicate the seasonal

variations of the wind speed data of the stations.

Two non-parametric statistics are in common use in trend studies: Mann-Kendalls (M-K) test and Spearmans test. Compared to the parametric tests, the non-parametric tests have been proved to provide higher statistical power in cases of non-normality of the distribution, and they are robust against outliers and missing

data (Turkes 1996, Turkes, 1999; Turkes et al, 2008, Zhihua et al, 2013). Furthermore, non-parametric tests

represent a measure of monotonic dependence whether linear or not (Davies, 1986; Rossi et al, 1992) in De Luis

et al (2000). In this work, Mann-Kendalls (M-K) rank correlation tests were chosen to detect significant trends.

4.2 The Mann-Kendall (M-K) Correlation Test

Within the M-K test, the data (1 , 2 , . , ) of time series as null hypothesis, Ho, are independent identically distributed random samples. Given n size data for n10, the M-K test statistic S is defined as follows (Rai et al, 2010; Zhihua et al, 2013):

= ( )

=+1

1

=1

(1)

Where xj and xj are the sequential data for the ith

and jth

terms, and j >i

=

1 , > 0, = 1 <

(2)

When S is a large positive number, later values exceed earlier values and upward trend is indicated.

When later values are less than earlier values, S is negative and downward trend results. Under the null

hypothesis of independent and randomly distributed random variables, when n10, the S statistic is approximately normally distributed, with zero mean and variance as follows in the absence of ties:

2 = 1 (2 + 5)

18 . . (3)

The value of S and 2 are used to compute the Z statistic, which follows a normal standardized distribution thus:

S 1, S> 0..(4)

0 , S = 0

S +1 , S < 0

The null hypothesis Ho that there is no trend is rejected when the absolute Z value computed by eqn (4)

is greater than the critical value at a chosen level of significance. Conversely, the alternative hypothesis H1 that the data follow a monotonic trend over time is accepted. The test statistic tau () is computed as

=

(1)/2 (5)

In this study, the Z value is tested at the 1% and 5% significance levels. The trend is upwards for

positive values of Z and downwards for negative values of Z. The test statistic (Kendalls tau b) has a range of -1 to +1, and is analogous to the correlation coefficient in regression analysis. The null hypothesis is rejected

when the tau b () is significantly different from zero. To test the trend significance, Z is computed and the cumulative probability for a standard normal distribution at /Z/ is found. For a two tailed test, the value of the

cumulative probability is multiplied by 2 to obtain the p value. If the p value is below a given level of

significance, the trend is significant.

4.3 Linear Regression

The Mann-Kendall tests described above detect nature and significance of the trends but do not provide

an estimate of their magnitudes. The slopes (magnitudes) of the trends were quantified by a linear regression

model of the form:

= + 6

Z =



Where y is the wind speed (in m/s), x is the number of months, m is the slope of the trend (in m/s per

month) indicating the detected change, and C is a regression coefficient (the intercept). When the slope m is

positive, it means that the wind speed has an upward trend and vice versa. The larger the absolute value of m,

the more obvious the variation trend is.

V. Results And Discussion Table 2 shows the wind speed characteristics of the stations. The coefficients of variation (CV) are high

for majority of the stations which indicates high wind speed variability across the country. A cursory look at the

table indicates that there is no discernible pattern of distribution of both the mean and the CV. Table 2 further

portrays Sokoto and Kano in the North West as the windiest stations with spectacular mean wind speeds of

7.471 m/s and 7.9 m/s respectively. The CV of 50.32% and 49.37% are outstanding for the north east stations of

Yola and Bauchi respectively. The monthly mean of the wind speed data are relatively low in the south western

cities of Oshogbo (2.966 m/s), Benin (3.481 m/s), Warri (3.23 m/s) and in the north central city of Lokoja (3.063

m/s). The variation of wind speed across the stations as observed in table 2 could be attributed to a number of

potential causes ranging from orographic, orogenic and topographic features. Roughness of the environment

surrounding the stations, variations in the height and position of anemometers, and atmospheric forcing

(atmospheric circulation) changes also produce substantial effects. Some studies (e.g Bichet et al, 2012) have

found that increasing the vegetation roughness length (caused by increasing vegetation) decreases the land wind

speed. Wind speed tend to be higher at well exposed sites than at stations in the vicinity of forests, hills,

mountains and other intervening structures such as high rise buildings. Suffice it to say that changes in measured

wind speed can result from both atmospheric and ground surface controls. The result observed here is expected

since the north belongs to the arid and semi-arid ecotypes while the south is dominated by mangrove, swamp

forests, tropical rainforests and guinea savanna tall grasslands.

The Mann-Kendalls test results presented in table 3 show the Kendalls tau b (coefficients of the time trends) for the individual stations along with the p values of the test statistic. The levels of statistical significance

of the time trend coefficients are also indicated. The extreme right columns of table 3 show the estimates of the

trend magnitudes in m/s per month, m/s per year and m/s per decade. The dominant trend in the time series of

the wind speeds is the decline over the periods considered here. 11 stations (representing 55%) show downward

trends out of which 8 stations (representing 40% of the stations) show decreasing trend significant at the 1%

level. These are Sokoto, Kaduna, Bauchi, Yola, Oshogbo, Benin, Lokoja and Port Harcourt. 9 stations

(representing 45%) show upward trend out of which 7 stations (representing 35%) show significant upward

trend at the 1% level. These are Kano, Maiduguri, Ilorin, Ikeja, Enugu, Calabar and Makurdi. Owerri and Warri

show upward trends that are not significant while Yelwa, Ibadan and Ogoja show non-significant downward

trends at the chosen levels of significance.

Table 4 is the Person Product Moment Correlation matrix of monthly mean wind speeds. Their

significance levels are also shown to give a general indication of coincidence between stations and index time

series. The inter station spatial coherence of the monthly mean wind speed is objectively quantified by using the

station to station correlation coefficients. Some of the anemometer station pairs show negative correlations

significant at the 1% and 5% levels. Other station pairs show positive correlations significant at the 1% and 5%

levels. There are equally other station pairs that show positive and negative correlations that are not significant

at the chosen levels of significance. There is no discernible pattern in the correlation coefficients among the

pairs of anemometer stations.

Table 2: Descriptive statistics for wind speed Stations N Minimum Maximum Mean Std. Deviation Range C.V (%)

Yelwa 540 0.0 8.0 3.524 1.4164 8.0 40.19

Sokoto 540 2.4 12.9 7.471 1.9474 10.5 26.07

Kaduna 504 2.4 10.9 5.285 1.4716 8.5 27.84

Kano 576 2.0 15.0 7.900 2.4256 13.0 30.7

Bauchi 492 0.0 10.9 4.621 2.2814 10.9 49.37

Maiduguri 564 1.2 9.1 5.227 1.6468 7.9 31.51

Ilorin 588 1.0 8.6 4.411 1.4821 7.6 33.6

Yola 588 0.5 11.0 3.993 2.0091 10.5 50.32

Ikeja 744 0.3 10.5 4.635 1.6960 10.2 36.59

Ibadan 564 0.2 8.6 4.060 1.2619 8.4 31.08

Oshogbo 552 0.0 5.6 2.966 1.1029 5.6 37.18

Benin 552 1.1 8.2 3.481 0.9266 7.1 26.62

Warri 384 0.5 6.8 3.231 0.6419 6.3 19.87

Lokoja 576 0.3 6.2 3.063 1.0555 5.9 34.46

Port Harcourt 636 0.0 7.0 3.633 0.9971 7.0 27.45

Owerri 432 1.8 8.1 3.455 0.7840 6.3 22.69

Enugu 588 0.0 11.0 5.282 1.4183 11.0 26.85



Calabar 624 0.0 8.3 3.871 1.1474 8.3 29.64

Makurdi 540 1.3 9.2 4.781 1.4973 7.9 31.32

Ogoja 348 1.2 9.8 3.633 1.1195 8.6 30.81

Table 3: Mann-Kendalls test results and estimates of trend magnitudes of the wind speed. Stations Kendalls tau b p value m/s/month m/s/year m/s/decade

Yelwa -0.040 0.175 0.0013 0.0156 0.156

Sokoto -0.094** 0.001 0.0012 0.0144 0.144

Kaduna -0.095** 0.002 0.002 0.024 0.24

Kano 0.354** 0.000 0.0048 0.0576 0.576

Bauchi -0.382** 0.000 0.0009 0.0108 0.108

Maiduguri 0.178** 0.000 0.0014 0.0168 0.168

Ilorin 0.295** 0.000 0.0018 0.0216 0.216

Yola -0.438** 0.000 0.0058 0.0696 0.696

Ikeja 0.121** 0.000 0.0002 0.0024 0.024

Ibadan -0.043 0.134 0.0028 0.0336 0.336

Oshogbo -0.162** 0.000 0.005 0.06 0.6

Benin -0.097** 0.001 0.0001 0.0012 0.012

Warri 0.003 0.926 0.0003 0.0036 0.036

Lokoja -0.209** 0.000 0.001 0.012 0.12

Port Harcourt -0.271** 0.000 0.0015 0.018 0.18

Owerri 0.025 0.443 0.00039 0.00468 0.0468

Enugu 0.193** 0.000 0.0011 0.0132 0.132

Calabar 0.317** 0.000 0.0022 0.0264 0.264

Makurdi 0.259** 0.000 0.0018 0.0216 0.216

Ogoja -0.052 0.153 0.00042 0.00504 0.0504

** Kendalls tau b is significant at the 0.01 level (two-tailed).

Table 4 Correlation coefficients for Wind Speed across the stations

The observed changes over time in the measured wind speed data can result from both atmospheric

circulation changes and ground surface variations. Ramanathan et al, (2001) noted that aerosol emissions,

greenhouse gas concentrations, sea surface temperatures, can affect the atmospheric circulation and stability

thereby wind speeds. Bichet et al, (2012) observed that sea-induced circulation changes have a regional

character and can decrease or increase the wind speeds, whereas in contrast, higher aerosol concentrations

appear to generally reduce the land and ocean wind speeds. This response could be linked to the role of

atmospheric aerosols upon the stratification of the atmosphere. Whereas increasing aerosols emissions cool the

surface, carbonaceous aerosols also warm the aerosol layer in the troposphere. This would increase the

atmospheric density gradient between the surface and the troposphere, and thus reduce the rate of atmospheric

circulation (Ramanathan et al, 2005). Suffice it to say that the observed dominant downward trends in wind

speed could be linked to climate variability and change.



The time series plots with the trend lines (not shown) indicate that the trend lines uphold the Kendalls test result with respect to the direction of trends in the stations.

The seasonal variations of monthly mean wind speed are presented in Bar charts in Figs 2a t. The figures show that for the north western cities of Yelwa, Kaduna, Sokoto and Kano, there is a bimodal maxima.

These occur in Winter (Dec, Jan, Feb) and early Summer (June). In this region, the most calm periods are

observed from late summer (August) to mid autumn (October). Yelwa station is exceptional here in that the

wind speed reaches its crescendo between mid spring (April) to early summer (June), and its lowest in late

autumn (November).

The north eastern cities of Bauchi, Maiduguri and Yola record their maxima in mid spring (April) and

their minima in autumn (Sept, Oct, and Nov). However, Maiduguri deviates slightly from this pattern in that it

records its maxima in early spring (March) and early summer (June). The north central cities (Lokoja, Ilorin,

and Makurdi) have their windiest period in spring (March, April & May) and their most calm period in autumn

(Sept, Oct, Nov), similar to the situation in the north eastern zone. The south western cities (Ibadan, Oshogbo

and Ikeja) recorded double maxima which are observed in spring (March and April) and summer (July and

August).



Their minima are observed in late autumn (November). For cities in the south eastern geographical

zones (Enugu, Owerri and Ogoja), maxima are observed in spring (March and April) and the minima occurred

in late autumn (November). For the core southern cities of Benin, Port Harcourt, Warri and Calabar, double

maxima are observed in spring (March & April) and late summer (August). The calm periods are witnessed in

late autumn (November). From the foregoing, it is clear that the spring season (March and April in most cases)

dominates the periods of high wind speed while the autumn (November precisely) dominates the periods of

calm across Nigeria.

The result of this research is in complete agreement with that of Ogolo and Adeyemi (2009) that

observed declining trends in wind speed for Ibadan using the M-K test. However, the results do not partially

agree with that of Ewona and Udo (2008) that observed decreasing trends in wind speed for Calabar. The

variation in result is perhaps, due to differences in record length of the data used. The result of this work is

consistent with other regional and international studies where the results indicate dominant declining trends in

the wind speed data (Tuller, 2004;Abhishek et al, 2010; Bichet et al, 2012; Ko et al, 2010b; Mc Vicar et al,

2010; Troccoli et al, 2012;). Presence of decreasing, increasing and random trends across the stations as shown

in the M-K test show that the direction of trends and variations are location dependent as observed by some

studies (e.g Kumar and Philip, 2010; Wu and Mok, 2013 and Koet al, 2010a).

VI. Conclusion It has been the main objective of this study to examine the trends and variations of measured wind

speed data at 20 land anemometer stations spread across Nigeria and to quantify the statistical significance of

the trends. The coefficients of variations (CV) are high, ranging from 19.87% to 50.32%. The northern part of

the country tends to show higher CV and monthly mean daily wind speeds. The inter station spatial coherence of

the monthly mean daily wind speed as quantified by the Pearson Product Moment Correlation Coefficients

indicate some negative and positive correlations significant at the 0.01 and 0.05 levels. The Mann-Kendalls test results show a dominant decreasing trend in the time series of the period considered in the study, of which most

of them are significant at the 0.01 level. Nevertheless, there are also upward trends of which some are

significant at the 0.01 level. It is clear from the seasonal variation Bar charts that the spring period (March and

April precisely) is the windiest period while autumn (precisely November) dominates the period of calms across

Nigeria.

The results of the M-K tests portray the fact that surface wind pattern could be an alternative to surface

air temperature and precipitation as an indicator of climate variability and change. Variations in wind speed

pattern may provide a critical context for climate change research and a platform for forecasting and modeling

of wind speed regimes under the global climate change scenarios. The knowledge of contemporary wind climate

data and its historical trends can be useful to various agencies and industries. The results presented here have

implications for the air quality management attempts in these regions as decreasing wind speed would affect

ozone and aerosol distribution rates and patterns. Consequently, this may signal a need to make significant

changes to their air pollution mitigation strategies for effective results. This is necessary because downward

trends in wind speed may imply poorer ventilation of pollutants from these areas which could constitute serious

health-related problems especially for people with heart-related diseases such as asthmatic patients.

References [1]. Abhishek, A., Lee, J.Y., Keener, T.C. and Yang, Y.J. (2010). Long-term wind speed variations for Three Midwestern U.S cities. J.

Air & Waste Manage. Assoc. 60, 1057 1064, doi:10.3155/1047 3289.60.1057. [2]. Abiodun, B.J., Salami, A.T. and Tadross, M. (2011).Climate Change Scenarios for Nigeria: Understanding the Bio-Physical

Impacts. Climate Systems Analysis Group, Cape Town, for Building Nigerias Response to Climate Change (BNRCC) Project, Ibadan, Nigeria.

[3]. Amadi, S. O., Udo, S. O. and Ewona, I. O. (2014). Trends in Monthly Mean Minimum and Maximum Temperature Data over Nigeria for the Period 1950-2012. International Research Journal of Pure and Applied Physics, 2(4), 1-27.

[4]. Bichet, A., Wild, M., Folini, D. and Schar, C. (2012). Causes for decadal variations of wind speed over land: sensitivity studies with a global climate model. Geophysical Research Letters, 39, L11701, doi:10.1029/ 2012GL051685.

[5]. Cabezudo, B., Recio, M., Sanchez-Laulhe, J., Del Mar Trigo, M., Toro, F.J. and Polvorinos, F. (1997). Atmospheric transportation of Marihuana Pollen from North Africa to the southwest of Europe.Atmos. Environ, 31, 3323 3328.

[6]. Cardone, V.J., Greenwood, J.G. and Cane, M.A. (1990).On Trends in Historical Marine Wind Data.Journal of Climate.3,113 127. [7]. Davis, J.C. (1986). Statistics and data analysis in Geology, Wiley, New York. [8]. De Luis, M, Raventos, J, Gonzalez-Hidalgo, J.C, Sanchez, J.R and Cortina, J, (2000). Spatial analysis of rain-fall trends in the

region of Valencia (East Spain), International Journal of Climatology, 20, 1451 1469. [9]. Droppo, J.G. and Napier, B.A. (2008). Wind direction bias in generating wind roses and conducting sector-based air dispersion

modeling. J. Air & Waste Manage. Assoc., 58,913 918, doi:10.3155/1047 3289.58.7.913. [10]. Ewona, I.O. and Udo, S.O. (2008). Trend studies of some meteorological parameters in Calabar, Nigeria. Nigerian Journal of

Physics, 20(2): 283 289. [11]. Helsel, D.R. and Hirsch, R.M. (1992).Statistical methods in water resources. Studies in Environmental Science, 49, New York:

Elsevier. [12]. Holzer, M. and Boer, G.J. (2001).Simulated Changes in atmospheric transport climate.J. Clim, 14, 4398 4420.



[13]. Hosking, J.R.M. and Wallis, J.R. (1997).Regional frequency analysis: an approach based on L-moments. Cambridge University Press, Cambridge.

[14]. Hwang, M.K., Kim, Y.K., Oh, I.B., Lee, H.W. and Kim, C.H. (2007).Identification and Interpretation of representative ozone distributions in association with the sea breeze from different synoptic winds over the coastal urban area in Korea.J.Air& Waste

Manage. Assoc, 57, 1480 1488, doi:10.3155/1047-3289. 57.12.1480. [15]. Jacob, D.J. and Winner, D.A. (2009).Effect of Climate Change on air quality.Atmospheric environment,43,51-63.

Doi:10.1016/j.atmosenv.2008.09.051.

[16]. Jungo, P., Goytette, S. and Beniston, M. (2002).Daily wind gust speed probabilities over Switzerland according to three types of synoptic circulation.Int.J.Climatol, 22, 485 499, doi:10.1002/joc.741.

[17]. Karabulut, M., Gurbuz, M. and Korkmaz, H. (2008). Precipitation and temperature trend analyses in Samsun. J. Int. Environmental Application and Science, 3(5), 399-408.

[18]. Karabulut, M., Demirci, A. and Kora, F. (2012).Analysis of spatially distributed annual, seasonal and monthly temperatures in Istanbul from 1975 to 2006.World Applied Sciences Journal, 12(10), 1662-1675.

[19]. Ko, K., Kim, K. and Huh, J. (2010a). Characteristics of wind variations on Jeju Island, Korea.International Journal of Energy Research, 34, 36-45, doi:10.1002/er.1554.

[20]. Ko, K., Kim, K. and Huh, J. (2010b). Variations of wind speed in time on Jeju Island, Korea. Energy 35:3381 3387. [21]. Kumar, V.S. and Philip, C.S. (2010). Variations in long term wind speed during different decades in Arabian Sea and Bay of

Bengal. J.Earth Syst. Sci, 119(5), 639 653. [22]. Mc Vicar, T.R. and Roderick, M.I. (2010). Atmospheric science: winds of change. Nat. Geosci., 3,747748, doi: 10.1038/ngeo1002. [23]. Mc Vicar, T.R., Van Niel, T.G., Roderick, M.L., Li, L.T., Mo, X.G., Zimmermann, N.E. and Schmatz, D.R. (2010). Observational

evidence from two mountainous regions that near-surface wind speeds are declining more rapidly at higher elevations than lower elevations: 1960 2006. Geophysical Research Letters, 37, 10.1029/2009GL042255.

[24]. National Nonpoint Source Monitoring Program (2011).Statistical analysis for Monotonic trends.Technotes 6, November 2011. [25]. Ngongondo, C., Xu, C. and Gottschalk, L. (2011). Evaluation of spatial and temporal characteristics of rainfall in Malawi: a case of

data scarce region. Theor. Appl. Climatol, 106, 79-93, doi:10:1007/S00704-011-0413-0.

[26]. Ogolo, E. O. and Adeyemi, B. (2009).Variations and trends of some meteorological parameters at Ibadan, Nigeria.The Pacific Journal of Sc. & Tech, 10(2), 981-987.

[27]. Ojo, O. (1977). The Climates of West Africa, Ibadan, Heinemann. [28]. ONeal, M., Nearing, M., Vining, R., Southworth, J. and Pfelfer, R. (2005). Climate change impacts on soil erosion in Mid West

United States with changes in crop management. Cantena, 61,165-184:doi:10.1016/j.cantena.2005.03.003. [29]. Rai, R.K, Upadhyay, A. and Ojha, C.S.P. (2010). Temporal variability of climatic parameters of Yamuna River Basin: Spatial

analysis of persistence, trend and periodicity. The Open Hydrology Journal, 4,184 210. [30]. Ramanathan, V and Feng, Y. (2009). Air pollution, greenhouse gases and climate change: Global and regional perspectives.

Atmospheric Environment, 43,37-50. Doi:10.1016/j.atmosenv.2008.09.063.

[31]. Ramanathan, V., Crutzen, P.J., Kiehl, J.T. and Rosenfeld, D. (2001).Aerosols, climate, and the hydrological cycle.Science, 294, 2119 2124, doi:10.1126/ science. 1064034.

[32]. Ramanathan, V., Chung, C., Kim, D., Bettge, T., Buja, L., Kiehl, J.T., Washington, W.M., Fu, Q., Sikka, D.R. and Wild M (2005). Atmospheric brown clouds: Impacts on south Asian climate and hydrological cycle. Proc. Natl. Acad. Sci. U.S.A, 102(15), 5326 5333, doi: 10.1073/pnas.0500656102.

[33]. Rayner, D.P. (2007). Wind run changes: the dominant factor affecting pan evaporation trends in Australia. J. Clim, 20, 3379 3394; doi: 10.1175/JCL 14181.1.

[34]. Roderick, M.L., Rotstayn, L.D., Farquhar, G.D. and Hobbins, M.L. (2007).On the attribution of changing pan evaporation, Geophys. Res. Lett, 34, L17403, doi: 10.1029/2007GL031166.

[35]. Rossi, R., Mulla, D., Journel, A. and Franz, E. (1992).Geostatistical tool for modeling and interpreting ecological spatial dependence.Ecological monograph, 62, 277-314.

[36]. Shongwe, M.E, Landman, W.A. and Mason, S.J. (2006). Performance of recalibration systems for GCM forecasts for Southern Africa. Int. J. Climatol., 17:1567-1585.

[37]. Syners, R. (1990). On the statistical analysis of series of observations. WMO Technical Note 43, World Meteorological Organization, Geneva.

[38]. Troccoli, A., Muller, K., Coppin, P., Davy, R., Russell, C. and Hirsch, A.L. (2012). Long-term wind speed trends over Australia. Journal of Climate, 25, 170 183.

[39]. Tuller, S.E. (2004). Measured wind speed trends on the west coast of Canada. Int.J.Climatol. 24:1359 1374. [40]. Turkes, M. (1996).Spatial and temporal analysis of annual rainfall variations in Turkey.Int. J. Climatol., 16: 1057 1076. [41]. Turkes, M. (1999).Vulnerability of Turkey to desertification with respect to precipitation and aridity conditions.Tr. J. of

Engineering and Environmental Sciences, 23: 363 380. [42]. Turkes, M., Koc, T. and Savis, F. (2008).Spatio-temporal variability of precipitation total series over Turkey.Int. J. Climatol.

Doi:10.1002/joc.1768.

[43]. Viles, H.A. and Goudie, A.S. (2003).International, decadal and multidecadal scale climate variability and geomorphology.Earth Sci. Rev, 61, 105 131; doi:10.1016/S0012 8252 (02) 00113 7.

[44]. Wu, M.C. and Mok, H.Y. (2013). Regional and seasonal variations of the characteristics of Gust factor in Hong Kong and recent observed long term trend. Journal of Civil Engineering and Science, 2(1), 1 6.

[45]. Wu, C.F., Chen, C.H., Chang, S.Y., Chang, P.E., Shie, R.H., Sung, L.Y., Yang, J.C. and Su, J.W. (2008). Developing and evaluating techniques for localizing pollutant emission sources with open-path Fourier Transform Infrared measurements and wind

data. J. Air Waste Manage. Assoc., 58,1360 1369, doi:10.3155/1047 3289.58.10.1360. [46]. Zhihua, Z, Rui, L., Hui, Q., Jie, C. and Xuedu, Z. (2013).Analysis of rainfall variation over the past 55 years in Guyuan City,

China.Journal of Environmental Science, Computer Science and Engineering Technology, 2(3), 640 649.

analysis of trends and variations of monthly mean wind speed data in nigeria

Documents